A. Baumberg and D. Hogg. An efficient method for contour tracking using active shape models. In Proc. IEEE Workshop on Motion of Non-Rigid and Articulated Objects, pages 194-- 199, 1994.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....sufficient image resolution is available and vehicle sizes do not greatly exceed the priors on object size used in object initialisation. 1 Introduction In recent years there has been much work on the tracking of moving objects within a scene. Systems developed for such tasks as people tracking [16, 1, 8, 6], face tracking [4, 9] and vehicle tracking [3, 13, 5] have come in many shapes or sizes, but may be broadly divided into explicit model based methods (where a fixed, detailed model of object characteristics is built e.g. 4] or implicit model based methods (where more general object, or scene, ....

....framerate per pixel (a integer division of the incoming image stream framerate) Stable pixels (i.e. pixels that fit the background distribution over a number of frames) are updated with a lower framerate than pixels that have been recently classified as foreground. The framerate is given by: Z [1]Q a6 cb ZedEf ]1[ hg i P j 2 where i P j lk iWm Tn ZedEf ]1[ hg iWm n 2 otherwise (7) Where: Ze[T]Q a = Update rate ZedEf ]1[ Input Framerate i P j = No. of consecutive background frames i P j = Threshold on frame rate division (typically 25 frames 1 ....

A. Baumberg and D. Hogg. An efficient method for contour tracking using active shape models. In Proc. IEEE Workshop on Motion of Non-Rigid and Articulated Objects, pages 194--199, 1994.

....of likely object motion and deformation over one time step allow objects to be tracked robustly through a scene. This has been achieved successfully using a Kalman filter framework [1] for example in fitting a deformable model of human shape to an image in order to track over image sequences [2]. Whilst such a framework is suitable for simple situations where the probability density function (pdf) of the input data can be represented as a single Gaussian distribution, it cannot support more than one simultaneous hypothesis. An alternative approach is that of the condensation algorithm ....

A.M. Baumberg and D.C. Hogg. An efficient method for contour tracking using active shape models. In Proc. IEEE Workshop on Motion of Non-rigid and Ariculated Objects, pages 194--199, Austin, Texas, November 1994.

....of human shape in isolated static images taken from a motion sequence; in this study we discuss an individual s shape deformation during motion. The method discussed is part of a larger system designed to track moving pedestrians, a problem that has been the subject of much research [4, 5, 6]. We show that the periodic nature of human walking is clearly discernible from the deformation pattern, and that individual humans have a specific temporal pattern. The paper is divided into 6 sections. Section 2 discusses the use of active contour models for detecting walking humans. Section 3 ....

. Baumberg A. M. & Hogg D. C. [1994]. An efficient method for contour tracking using active shape models. University of Leeds School of Computer Studies Research Report Series, Report 94.11.

....feasibility. As to the former, such a system could easily find its way into commercial applications, such as video conferencing, automatic surveillance, and distance learning. As to the latter, recent research has shown great progress in developing systems that are capable of performing this task [1, 4, 6, 7, 9, 10, 11, 13, 14, 15, 16, 17]. Despite the recent progress, however, at least two difficulties continue to impede the development of a robust, reliable tracker. First, many trackers assume that the subject does not perform any out of plane rotation (rotation about an axis parallel to the image plane) and fail when this ....

.... Gamma y t Gamma2 oe p t = oe t Gamma1 : Somewhat surprisingly, this simple prediction scheme greatly improves the behavior of the tracker because it removes any restriction on the maximum lateral velocity of the subject only the amount of acceleration is limited. A Kalman filter [1, 3] might improve results even more but has not been tried. 2 Since the gradients around the chin tend to be small, the particular choice of aspect ratio seems unimportant. Others [5] have advocated using the golden ratio (approximately 1.6) 3 We have not found it necessary to predict the size. ....

[Article contains additional citation context not shown here]

A. M. Baumberg and D. C. Hogg. An efficient method for contour tracking using active shape models. In Proceedings of the IEEE Workshop on Motion of Non-Rigid and Articulated Objects, pages 194--199, 1994.

....ior. As a result, the problem of determining the context for the interaction is the most critical sensing task required by the interface. An preliminary report on this work appears in [25] The second body of related work concerns algorithms for tracking human motion using video images [15, 12, 17, 4, 18, 26, 2]. Our color and motion blob algorithms are most closely related to those of Wren et al. 26] which are employed in the Alive system. The color histogram representation for blobs [20] that we employ is more descriptive than a single color blob model and therefore more appropriate to our task of ....

A. Baumberg and D. Hogg. An efficient method for contour tracking using active shape models. In J. Aggarwal and T. Huang, editors, Proc. of Workshop on Motion of Non-Rigid and Articulated Objects, pages 194--199, Austin, Texas, 1994. IEEE Computer Society Press.

....camera calibration and find the 3D position of the blobs (hands and face) In the end some applications which have used the P SP finder are presented. Comments: 99] gives a better overview of the Pfinder algorithm. Title: An Efficient Method for Contour Tracking using Active Shape Models [8] Author(s) A.M. Baumberg and D.C. Hogg Location: University of Leeds, UK Year: 1994 Published: School of computer studies research report series, Report 94.11 Type: Paper Key words: Active shape models, contour tracking, B spline, Principal components, Kalman filter and Sobel based edge ....

A.M. Baumberg and D.C. Hogg. An Efficient Method for Contour Tracking using Active Shape Models. Technical Report 94.11, University of Leeds, UK, 1994.

.... Sklansky [132] 1991 Wang et al. 142] 1991 Yamamoto and Koshikawa [152] 1992 Kepple [86] 1992 Lee and Chen [91] 1992 Luo et al. 102] 1992 Wang et al. 143] 1993 Kameda et al. 84] 1994 Aggarwal et al. 2] 1994 Baumberg and Hogg [13] 1994 Bharatkumar et al. 16] 1994 Darrell et al. 40] 1994 Gu et al. 59] 1994 Guo et al. 61] 1994 Niyogi and Adelson [117] 1994 Polana and Nelson [126] 1994 Rossi and Bozzoli [129] 1994 Schneider and Bekker [130] Table 3.1: ....

....These are described in chapter 7. 5.2 Generating User Models The result of a model based approach is usually dependent on how good the user fits the human model in the system. Some systems use a general model which is an average of standard humans, see e.g. Rohr [128] and Baumberg and Hogg [13]. Other systems measure the current user and generate a model based on these data. This can be done both off line and on line. In the work by Lerasle et al. 95] an off line example is given. They use 34 cross section RMI scans of a leg to generate a personalized model. In the work by Gu et al. ....

[Article contains additional citation context not shown here]

A.M. Baumberg and D.C. Hogg. An Efficient Method for Contour Tracking using Active Shape Models. Technical Report 94.11, University of Leeds, UK, 1994.

....of a known motion pattern. The result of a model based approach is usually dependent on how well the subject fits the human model in the system, i.e. the complexity of the model. Some 14 THOMAS B. MOESLUND AND ERIK GRANUM systems use a general model which is an average of many individuals [9]. Others measure the current subject and generate a model based on these data. This may be done off line [49] 85] or on line as by Wren et al. 141] In the latter, analysis of the subject s initial pose is carried out to build an initial model which is refined as more information becomes ....

....of the curve. This type of active contours is also called snakes and works well when the changes in the structure of the object is unknown. If a shape model is used the active contour is known as a deformable template [13] Active contours may be used to extract the entire outline of the subject [9] or to extract individual body parts [74] A related approach is to extract the silhouette instead of the contour. Rigoll et al. 120] use a stochastic approach to silhouette extraction. They use pseudo 2D Hidden Markov Models (HMM) nested one dimensional HMMs) to extract the silhouette in a ....

[Article contains additional citation context not shown here]

A.M. Baumberg and D.C. Hogg. An Efficient Method for Contour Tracking using Active Shape Models. In Workshop on Motion of Non-Rigid and Articulated Objects, pages 2--14, Austin, Texas, USA, 1994.

....Polana and Nelson [22] refer to getting your man without finding his body parts . Models for human action are described in statistical terms based on low level features. Foreground regions are typically obtained by skin color detection or background subtraction from which features based on shape [2] [6] 12] texture [7] 8] 21] or motion [4] 9] 22] are extracted. In some cases [12] 21] the requirement of a separate foreground segmentation is relaxed by the employment of window search procedures. The second approach uses explicit a priori knowledge of how the human body (or hand) ....

A. Baumberg and D. Hogg. An efficient method for contour tracking using active shape models. In IEEE Workshop on Motion of Non-Rigid and Articulated Objects, pages 194--199, Austin, 1994.

....1996, c fl IEEE96 2 in the images. These methods typically attempt to match the projection of a detailed articulated 3D body model to the edge data [9, 10, 16] A number of authors have extended active contour models to model articulated motion [6, 12, 13, 20] For example, Baumberg and Hogg [2] track the outline of a moving body using a modal based flexible shape model which captures the considerable outline variations in the human silhouette during movement. Stickfigure models of humans have also been matched to image data [1, 14] These methods are typically only applied to humans ....

A.M. Baumberg and D.C. Hogg, "An Efficient Method for ContourTracking using Active Shape Models," IEEE Workshop on Motion of Non-rigid and Articulated Objects, 1994, 194-199.

....drawbacks. Features need to be present at all times (no 8 occlusions) At initialization, a good initial estimate must be available for the method to converge properly. And finally, the chosen parametrization might include states which have implausible physical interpretations. Baumberg and Hogg [8] apply Active Shape Models to the tracking of pedestrians. They use a somewhat different shape representation, based on B splines; see Figure 3. By assuming a stationary camera, tracking is initialized on the foreground region; the latter is obtained by background subtraction. Spatio temporal ....

....work by Heap and Hogg [31] involves an example based approach 16 to articulated pose recovery. Their method involves a principal component analysis of 3 D positional (hand) data and allows shape deformations of a tracked object. This method was mentioned earlier in the 2 D context; see Section 3 [8]. 5.3 Feature correspondence A variety of features can be used to establish correspondence between model and image remains, from low level to high level. Using high level features (e.g. joint locations) simplifies pose recovery but places a greater burden on segmentation. Approaches [17] 37] ....

[Article contains additional citation context not shown here]

A. Baumberg and D. Hogg. An efficient method for contour tracking using active shape models. In Proc. of IEEE Workshop on Motion of Non-Rigid and Articulated Objects, pages 194--199, Austin, 1994.

....and Labelling This part is concerned with following the corresponding lights from one frame to the next and drawing the locus of the trajectories of each marker. Different tracking algorithms are mentioned in literature, varying in their level of complexity and accuracy see for example [1] and [4] Our algorithm uses a simple, yet reliable technique for tracking. The procedure assumes smooth motion. In the first frame having a window of fixed size is placed around each marker position, the Euclidean distance of each pair of coordinates of two successive frames is calculated for all ....

A. M. Baumberg and D. C. Hogg. An efficient method for contour tracking using active shape models. Technical Report 94.11, University of Leeds - School of Computer Studies, Division of Artificial

....drawbacks. Features need to be present at all times (no occlusions) At initialization, a good initial estimate must be available for the method to converge properly. And finally, the chosen parameterization might include states which have implausible physical interpretations. Baumberg and Hogg [8] applied active shape models to the tracking of pedestrians. They used a somewhat different shape representation, based on B splines; see Fig. 3. By assuming a stationary camera, tracking is initialized on the foreground region; the latter is obtained by background subtraction. Spatiotemporal ....

....shifting windows of various sizes over the image, normalizing for gradient energy within the window,and determining FIG. 3. Principal component analysis on a data set of pedestrians represented by B splines; shown is the shape variation along the principal component (from Baumberg and Hogg [8], c # 1994 IEEE) 86 D. M. GAVRILA FIG. 4. Principal component analysis on a data set of pedestrians represented by images of size 30 by 50 pixels; shown are eigenvectors 0 (mean) 1, 2, and 25, in order of decreasing eigenvalues (from Franke et al. 23] the distance between the ....

[Article contains additional citation context not shown here]

A. Baumberg and D. Hogg, An efficient method for contour tracking using active shape models, in Proc. of IEEE Workshop on Motion of Non-Rigid and Articulated Objects, Austin, 1994, pp. 194--199.

....for recognition. 2.4 Implementation 2.4.1 Tracking Following the dots from one frame to the next and drawing the locus of the trajectories is the next stage. Different tracking algorithms are mentioned in literature, varying in their level of complexity and accuracy see for example [4] and [1]. Our algorithm uses a simple, yet reliable technique for tracking. The algorithm calculates the Manhattan distance of each pair of co ordinates (of two successive frames) and then applies a labelling procedure to label the different co ordinates in all frames based on a priori knowledge about ....

A.M Baumberg and D.C. Hogg. An efficient method for contour tracking using active shape models. Technical Report 94.11, University of Leeds - School of Computer Studies, Division of Artificial Intelliegence, April 1994.

....represented as an n vertex polygon. The n vertex polygons are created by placing points over the object boundaries either manually or semi automatically. Training is done by estimating the covariances of chord lengths between each pair of vertices in the polygons created. Later Baumberg and Hogg [2] used this idea to track contours by associating points with B splines rather than edges of a polygon. This method is dependent on the initial placement of the points of the polygon and due to this reason models only global deformations. Sclaroff and Pentland [32, 38] had proposed a method of ....

A. M. Baumberg and D. C. Hogg. An efficient method for contour tracking using active shape models. Technical Report 94.11, University of Leeds, School of Computer Studies Research Report Series, April 1994.

....recognition approaches in that it does not require any type of object or motion specific model. Other motion based recognition techniques [6, 7, 9, 10, 11, 12, 13] require a priori models of the underlying object and or the motion in the scene, although these models can potentially be learned [9, 15]. In contrast, periodicity is a universal motion characteristic that can be detected and described without knowledge of the underlying object and without reference to a previous instance of the motion. Therefore, the techniques in this chapter can be used to interpret image sequences of completely ....

A. M. Baumberg and D. C. Hogg, "An efficient method for contour tracking using active shape models," in Proc. Workshop on Motion of Non-Rigid and Articulated Objects, pp. 194--199, 1994.

No context found.

A. Baumberg and D. Hogg. An efficient method for contour tracking using active shape models. In Proc. IEEE Workshop on Motion of Non-Rigid and Articulated Objects, pages 194-- 199, 1994.

No context found.

A. Baumberg and D. Hogg, "An efficient method for contour tracking using active shape models," in Proc. of IEEE Workshop on Motion of Non-Rigid and Articulated Objects, (Austin), pp. 194--199, 1994.

No context found.

A. Baumberg and D. Hogg, "An efficient method for contour tracking using active shape models", Proc. of the Workshop on Motion of Nonrigid and Articulated Objects, 1994.

No context found.

A. M. Baumberg and D. C. Hogg. An efficient method for contour tracking using active shape models. IEEE Workshop on Motion of Non-rigid and Articulated Objects, Nov. 1994.

No context found.

A. Baumberg and D. Hogg. An efficient method for contour tracking using active shape models. In IEEE Workshop on Motion of Non-rigid and Articulated Objects, pages 194--199, November 1994.

No context found.

A. Baumberg and D. Hogg, "An efficient method for contour tracking using active shape models," in Proc. of IEEE Workshop on Motion of Non-rigid and Articulated Objects, pp. 194-199, IEEE Computer Society Press, Nov. 1994.

No context found.

A. Baumberg and D. Hogg, "An efficient method for contour tracking using active shape models", Proc. of the Workshop on Motion of Nonrigid and Articulated Objects, 1994.

No context found.

A. Baumberg and D. Hogg. An efficient method for contour tracking using active shape models. In Proc. of IEEE Workshop on Motion of Non-Rigid and Articulated Objects, pages 194--199, 1994.

No context found.

Baumberg, A. and Hogg, D. An efficient method for contour tracking using active shape models. In: Proceedings of the Workshop on Motion of Nonrigid and Articulated Objetcs. IEEE Computer Society, 1994.

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC